dc.contributor.author | Acer, Seher | en_US |
dc.contributor.author | Selvitopi, Oğuz | en_US |
dc.contributor.author | Aykanat, Cevdet | en_US |
dc.coverage.spatial | Santiago de Compostela, Spain | |
dc.date.accessioned | 2018-04-12T11:46:45Z | |
dc.date.available | 2018-04-12T11:46:45Z | |
dc.date.issued | 2017-08-09 | en_US |
dc.identifier.uri | http://hdl.handle.net/11693/37650 | |
dc.description | Date of Conference: 28 August - 1 September, 2017 | |
dc.description | Conference name: Euro-Par: European Conference on Parallel Processing - Euro-Par 2017: Parallel Processing
23rd International Conference on Parallel and Distributed Computing | |
dc.description.abstract | The scalability of sparse matrix-vector multiplication (SpMV) on distributed memory systems depends on multiple factors that involve different communication cost metrics. The irregular sparsity pattern of the coefficient matrix manifests itself as high bandwidth (total and/or maximum volume) and/or high latency (total and/or maximum message count) overhead. In this work, we propose a hypergraph partitioning model which combines two earlier models for one-dimensional partitioning, one addressing total and maximum volume, and the other one addressing total volume and total message count. Our model relies on the recursive bipartitioning paradigm and simultaneously addresses three cost metrics in a single partitioning phase in order to reduce volume and latency overheads. We demonstrate the validity of our model on a large dataset that contains more than 300 matrices. The results indicate that compared to the earlier models, our model significantly improves the scalability of SpMV. © 2017, Springer International Publishing AG. | en_US |
dc.language.iso | English | en_US |
dc.source.title | Euro-Par 2017: Parallel Processing - 23rd International Conference on Parallel and Distributed Computing | en_US |
dc.relation.isversionof | https://doi.org/10.1007/978-3-319-64203-1_45 | en_US |
dc.subject | Communication cost | en_US |
dc.subject | Hypergraph partitioning | en_US |
dc.subject | One-dimensional partitioning | en_US |
dc.subject | Sparse matrix-vector multiplication | en_US |
dc.title | Addressing volume and latency overheads in 1d-parallel sparse matrix-vector multiplication | en_US |
dc.type | Conference Paper | en_US |
dc.department | Department of Computer Engineering | en_US |
dc.citation.spage | 625 | en_US |
dc.citation.epage | 637 | en_US |
dc.identifier.doi | 10.1007/978-3-319-64203-1_45 | en_US |
dc.publisher | Springer | en_US |
dc.contributor.bilkentauthor | Aykanat, Cevdet | |